Calcrete-Capped Megaflood Gravel - George, WA
George Gravel Pit. Location is NW of George, WA. View to the SW.
George, WA is is a small town located off I-90 northeast of Vantage. The exposure described here is an old gravel pit near town. Take Exit 149 onto Hwy 281/Quincy/Wenatchee. Turn west onto Hwy 281 (opposite side of freeway from gas station). Follow Hwy 281 a short distance and turn L onto Beverly-Burke Rd, then R onto Rd 1 NW. A square green metal-sided building with a small parking lot is your destination: The George Landfill, which at some point in the past was called the George Drop Box, is mostly abandoned, a property of Grant County Public Works (office in Ephrata). Please respect any signage they have posted. I've not had any issues accessing the pit. Exposures are along the east and south walls. GPS: 47.100418, -119.870179.
Baker (1973), following Bretz et al. (1956) and Richmond et al. (1965), described the George deposit as "pre-Bull Lake flood gravels" mostly because of the thick cap of calcrete (a paleosol of considerable age). Richard Waitt refers to this site as "Burke", which is the nearest rail station sign. Calcrete this thick requires hundreds of thousands of years to form. Bull Lake is a somewhat antiquated term that refers to the penultimate glacial episode preserved in the central Rocky Mountains. The Last Glacial Period is known there as the Pinedale glaciation, which is called the Fraser in Washington and BC (and elsewhere). The Pacific Northwest sorely needs to formalize its own glacial-period terminology. Its high time we distinguish our record from that of the Rocky Mountains. "Fraser" is start, but the name just seems too Canadian for many American geologists to embrace. Me included. How about "Northwest glaciation"?
The east and south walls of the pit expose >4m of old flood gravels (pebble-cobble-boulder) with south- and east-dipping foresets capped by a meter-thick, blocky to platy, fully-plugged calcrete (> Stage III). The bouldery gravel contains weathered and unweathered basalt clasts, light-colored granite, minor quartzite, and a few chunks of Vantage sandstone (a interbed in the basalt). Crumbly green-white-yellow-gray opal comes from rusty pillow-palagonite complexes in the basalt. A few compact, unfractured chunks of opal suitable for flintknapping projectile points can be found; look for the light gray stuff. Other nearby locations where pillow basalts and palagonite are found: Silica Rd offramp of I-90 and the intersection of Rd 9 SW and Rd Q SW.
Easy access. The gravel pit at George is an inactive and easily accessible quarry located a few steps from the pavement. The pit was first described by geologists many decades ago, but few details, photos, or sketches are found in their notes.
Meter-thick calcrete. Roughly a meter of calcrete overprints coarse flood gravels at George. This meter-thick unit seems to occur in many locations in the region.
Clasts exotic and mundane. Clasts of fresh Miocene Columbia River Basalt and associated chunks of green-yellow opal and rusty palagonite (derived from local pillow basalt complexes) are the dominant lithologies found in the gravels at George. Exotic lithologies include granite and minor quartzite in cobble sizes. Deeply-weathered clasts of basalt are also common.
How Many Floods?
Many believe there is more than one flood represented here, but a distinct sedimentological break between floods is difficult to tease out. A few meters below the calcrete ledge is a thinner, discontinuous, yet conspicuous second calcrete horizon (~15cm thick). It parallels the ledge in places and follows foresets in others, suggesting it is not a paleosol, but represents a somewhat complicated pathway followed for a time by groundwater. I struggled to find throughgoing erosional surfaces, fining-upward relationships, or other distinct breaks in the gravel pile. Let me know what you find.
Deeply Weathered Basalt
Either the deeply weathered basalt clasts in pre-Wisconsin flood deposits are:
a.) Rip-ups of already-weathered material dislodged from rocky footings by overland floods (weathering then transport)
b.) Fresh clasts dislodged by floods and weathered in place over a long period (transport then weathering)
The first option infers the period of weathering occurred during the Miocene and/or Pliocene (weathered bedrock). The second, demands all weathering took place during the Pleistocene (weathered gravel).
So how to determine which is correct? A few things come to mind.
- Spoil piles. The Quincy Chute is a major canal in the Columbia Basin Irrigation system. It crosses White Trail Rd/Rd 5 SW just west of the Colockum Golf Course. Voluminous spoil piles left by the dredge that created it contain abundant weathered basalt, calcrete, and unweathered basalt. The canal is >5 meters deep, excavated into bedrock and below the scabland surface. Its bottom lies well below any surficial deposits left behind by floods. The dredge material was in place prior to canal construction. Horizontal strata are exposed in the walls of the canal near the road. This indicates at least some weathering of the basalt took place while it was in place, that the weathered zone was fairly thick, and entire weathered zone at the site was not completely excavated by the floods.
- Outcrops. If the basalt clasts in a pre-Wisconsin flood gravel weathered together after deposition, then wouldn't many, if not most, bedrock knobs and buttes that never saw flooding (above flood trimlines) show similar depth of weathering? I don't see that. In fact, spheroidal weathering in the stack of CRB flows is hit and miss. The same flow in one location may be deeply weathered, but in another appear fresh. In high cliffs where numerous flows can be seen, is weathering systematically greater in the older flows? I don't see that either.
- Unweathered clasts. The occurance of fresh and weathered basalt clasts in a gravel deposit suggests that the gravel pile did not weather as a unit. The whole pile should be equally weathered. One clast should look like the next and the matrix should be reddened or otherwise show evidence of uniform diagenesis. If that's the case, then great - it's an old gravel that weathered after deposition. At George, both weathered and unweathered clasts occur together in the same gravel beds.
- Corestones. Another observation, perhaps related, is pre-rounded core stones. Roadcuts along Hwy 262 west of Hwy 17 and just upstream of Drumheller Channels contain "boulders" (corestones) that are still attached to large columns. Downstream, boulders liberated by the floods are found. Their round shape was not produced by tumbling along in a flood. Rather, in situ weathering did it.
I urge caution when interpreting the age of gravels containing weathered clasts. Early floods through the Channeled Scablands encountered not only loess, but broad swaths of deeply-weathered basalt exposed at the surface - a fractured, crumbling zone perhaps a meter or more thick. Non-basaltic clasts, libertated from pre-weathered footings or left behind by ancient streams to weather in the open, would also have been encountered. Rushing floodwaters easily quarried the material, entrained it, and deposited it along floodway tracts as bars. Fragile weathered basalt, like fragile mudstones from the Ringold and fragile sandstones from local interbeds, can survive some amount of transport. Start with a boulder, end with a cobble. To my eye, the presence of weathered clasts in a megaflood gravel may indicate great age, but corroborating evidence is needed (i.e., a thick paleosol cap, clear stratigraphic position, reversed magnetism, etc.).
Discontinuous lenses of reddish, silty-sandy sediment that contain abundant cemented root casts and burrows occur with the gravel just below the capping calcrete. Identical sediments are commonly associated a several flood deposit localities in the Othello-Saddle Mts-Paradise Flats area and appear to be a fine grained flood facies that attracted plants and burrowing critters later.
Calcretes Are Lowland Deposits
One of the most important, but least emphasized, aspects of calcretes in this part of Eastern Washington is their formation in lowland settings. Thick calcretes are closely associated with alluvial bottomland sediments, low-gradient alluvial fans, and sometimes ancient megaflood deposits. They developed near the paleo-water table, likely within the capillary fringe. People have written thoroughly on thinner, less distinctive caliche horizons in Palouse loess, but those are child's play. West of the Palouse, calcretes thicken considerably, becoming blanket-like layers that constitute a mappable, correlatable zone with regional extent. The calcretes are not a subunit of the upper Ringold, a commonly held belief of mapping crews in past decades (i.e., Grolier and Bingham, 1978). The so-called "capping calcretes" that today reside atop Yakima Fold Belt ridges formed in low-elevation, low-relief valley bottoms and low-angle alluvial fans. Later, they were elevated by faulting (Cooley, 2022).
Calcrete Type Sections West of the Palouse
The literature on the Palouse loess incorrectly identifies caliche/calcrete with loess hills and implies pedogenic carbonates are products of processes on upland hillslopes in windblown silt (see papers by Busacca and McDonald 1988-2012). However, field relationships in the Pasco Basin, on Pardise Flats, and at Saddle Mountains clearly show they were not. Calcrete developed fully (at least more fully) in the lower, drier region west of the Palouse and are far more interpretable there. Type sections for Eastern Washington's calcretes are best found west of the Palouse. The few thick calcretes that occur in loess-dominated sections are typically low in the section (earliest Palouse) and indicate periods when basinal conditions (climate, groundwater, sedimentation rate) expanded eastward, prior to the bulk of loess deposition and apparently prior to some tilting of the Palouse Slope.
Routes taken by the earliest scabland floods? Calcrete-capped flood gravels occur in eastern and western portions of the Channeled Scablands. A single floodpath cannot explain gravel occurrence in both places. Ancient floods appear to have converged on a Pliocene regional low at Othello (between the two arrows). Calcretes are thickest there, too. My speculative mapping on Bretz's 1932 basemap.
Did Ancient Floods Converge From Both East and West?
Floodwater that deposited the gravel at George flowed south down an ice-free Columbia River gorge (or Moses Coulee?), exited west into Quincy Basin near Babcock Ridge, and moved inland toward Drumheller Channels. Gravels and calcretes at White Trail (Cooley, 2022, Stops 3 and 4), Frenchman Hills Wasteway (George Neff's "Winchester" site in Baker, 1973), and Herman Railcut (Cooley, 2022, Stop 12) are likely correlative. This early route through Drumheller Channels contrasts with that of early flooding (and calcrete-capped gravels) in the eastern Scabland (SE of Ritzville) - at Harder Road and Marengo railcut (Baker et al., 1991). Evidence for at least one early flood is preserved there in the Telford tract. Floodwater escaped the Columbia near Spokane, moved south through loess hills toward paleo-Washtucna Coulee, and diverted west to scour the lower Eagle Lake surface, Esquatzel Coulee, and Ringold Coulee. Both flood routes appear to converge at a regional sag at modern day Othello. Drumheller Channels (Crab Creek), Lind Coulee (Crab Creek) appear to be long-lived, pre-scabland troughs (Pliocene alluvial valleys). It remains unclear how the timing of ancient flooding in the western Scabland (George, White Trail, Winchester) and eastern Scabland (Harder, Marengo) relate. Also unclear is their relation to paleosol-capped gravels in Pasco Basin to the south (White Bluffs Overlook, Old Maid Coulee).
Weathered basalt. A fractured and crumbling clast of weathered basalt stands apart from cobbles of fresh basalt. Are these rip-ups of a weathered bedrock or evidence of weathing of the gravel itself? Is the weathering post-Miocene or entirely Pleistocene?
Silty-sandy Lenses. Discontinuous lenses of reddish, silty-sandy sediment containing abundant root casts and cemented burrows occur just below the capping calcrete. Grainsize would suggest these are not flood deposits, but represent a post-flood loess bed or a proximal eolian sand sheet (proximal loess facies; reworked flood deposits). However, these sediments are commonly associated with flood sediments and occur in the same stratigraphic position as flood gravels. Here, it lies directly atop dark colored flood gravel and contains many rounded clasts.
Calcrete ribbons. Ribbons of CaCO3 with the right morphology suggest throughflow, inflation (vertical expansion during growth), and a larger proportion of fine grained sediment. Note how secondary carbonate scaffolds on the gravel (parent material) and appears to consume it. Fewer clasts are seen the higher you go in the calcrete unit, but many are still there. Also, this section is likely slightly thicker than it was when first deposited due to the inflative effects of carbonate growth.
You comments always welcome: email@example.com
Baker, 1973, GSA Special Paper 144
Baker et al., 1991, DNAG v.2 - Quaternary Non-glacial Geology, p. 215-250
Bretz, Smith, and Neff, 1956, GSA Bulletin v. 67
Cooley, S.W., 2022, Calcrete in the Channeled Scablands, Yakima Fold Belt, and Palouse Hills, NORTHWEST GEOLOGY - Journal of the Tobacco Root Geological Society, Geology of Wenatchee, Washington, v. 51, p.49-76
Grolier and Bingham, 1978, Washington DGER Bulletin 71